510 



THE AMERICAN NATURALIST 



[Vol. XLI 



being mostly small and immature shells of the same species as 

 those on the left. These small shells were of course much more 

 fragile than the mature ones. 



From the way in which the attempt to get the snails was begun, 

 it appears that when the shells are not broken the snails are dragged 

 out through the aperture. I did not see this done. To determine 

 whether it could be successfully accomplished, I seized an extended 

 snail with a heavy forceps and pulled upon it strongly. With 

 a steady pull the attachments to the shell slowly yielded and the 

 snail was removed almost entire. 



Psychology. 



I have described reels which were set at several of the burrows 

 to determine whether the animal that was moving the snails occu- 

 pied the burrows. The same reels were used to determine how 

 the occupied shells were distinguished from the empty ones. The 

 possible means that suggested themselves were the weight of the 

 snail, and the senses of touch, sight, and smell. 



Muscular Sense. — To learn whether weight was the criterion, 

 an empty shell was stuffed with sandy soil till it was about as 

 heavy as an occupied one. This, with an empty shell and an 

 occupied one, was placed near one of the burrows. Each shell 

 was tied to a reel, and all were placed at equally accessible points. 

 The occupied shell was drawn into the burrow at the time of the 

 first decided rise of temperature, while the other two were left 

 indefinitely. The experiment was repeated, but the occupied 

 shell was so placed that the shrew would have to go round the 

 empty and stuffed shells and under the reel in order to get it. 

 The occupied shell was again removed and the other two left. 

 The experiment was twice repeated at another burrow, with, the 

 same results. Evidently weight of shell is not the determining 

 feature. It seemed possible that the center of gravity might not 

 be at the same point in a <^tuffed shell as in an occupied one, and 

 that the shrew could detect this difference. Therefore the position 

 of the center of gravity in ;i ^tnllV-l and an occupied shell was 

 determined by balaneinu' on a knifr e<lg(> and by suspension; 

 it was found to be the same in the two shells. 



